Its very simple to see why. Lets take a fly as an example. Any living thing that is made up of multiple cells, starts off as a single cell. Think of each cell as a building block. This building block is a computing machine with a program embedded inside it to make the entire fly - legs, wings, eyes, etc. A fly may be made up thousands of such cells, each having a copy of the same program. Now each cell executes a portion of the whole program based on where it is located - a cell where the wing should be located executes the wing sub-program to make wings. A cell where the eyes should be, executes the eye sub-program to make eyes. If a cell where an eye should be, executes the leg sub-program, a leg will grow where an eye is. This is over simplified of course, but that's the basic idea. Cells need to know precisely where they are so they execute the right portion of the whole program.
So accuracy in determining position is key. For instance, a building block inside a growing 0.5 mm fly can determine its position to an accuracy better than .01mm. Some of our new cars have an accuracy of 50 feet (mine is 150 feet). So much for our GPS precision. Not only that. Nature's GPS system is very robust compared to our GPS system. All of us, who have driven in a city with high rise buildings know how unwise it is to rely purely on our car or phone GPS to navigate a new place. Nature's GPS algorithms can compensate for variations in its "GPS signals" - and still produce a fully functional fly.
So how do living things accomplish such high precision, compensating for variations in the positioning signals? Before talking of precision, how do cells know their position in the first place?
The "GPS technology" in cells at a basic level is very simple. Cells sense the concentration of select diffusing molecules (think a drop of ink spreading in water - that is diffusion), and turn on/off portions of the program within them. For instance imagine an oval shaped ball, containing a 1000 identical cells,each with the same program inside it - a program to make a fly. Now, imagine a high concentration of a molecule, that is only present on one side of this oval ball, and it diffuses into the rest of the ball (there are tricks nature has evolved to speed this transport too in some cases) creating a concentration gradient (i.e the concentration drops with distance from the source of the molecule). Now each cell has sensors that can bind to these diffusing molecules - so it can measure the concentration of the diffusing molecules - the more the concentration the greater the binding rate. When a cell senses a particular concentration, it may trigger execution of a particular portion of the program inside it. The varying concentration of a diffusing molecule can effectively trigger different sub-programs in cells located at different positions. So the "GPS technology" is as simple as reading the concentration of a diffusing molecule.
The elegance of this technology is its simplicity and robustness. Diffusion rates are only weakly sensitive to temperature and to the overall state of a cell. It is also surprisingly insensitive to randomly placed obstacles, even when densely packed. Also, even if the diffusing molecule transiently attaches to sticky molecules that are not sensors, these transient attachments only affect the rate at which the concentration gradient forms - it does not affect their steady state concentration gradient shape. Since I have extolled the virtues of diffusion and even compared its precision to our car GPS, it is only fair that I disclose a small detail - in case it looks as though I am suggesting anything along the lines of biomimicry. Diffusion is not the fastest way to know position, particularly when we are driving cars. It takes about an hour to establish a concentration gradient in a growing fly. But for the purposes of growth of a fly, this timeframe is adequate.
The elegance of this technology is its simplicity and robustness. Diffusion rates are only weakly sensitive to temperature and to the overall state of a cell. It is also surprisingly insensitive to randomly placed obstacles, even when densely packed. Also, even if the diffusing molecule transiently attaches to sticky molecules that are not sensors, these transient attachments only affect the rate at which the concentration gradient forms - it does not affect their steady state concentration gradient shape. Since I have extolled the virtues of diffusion and even compared its precision to our car GPS, it is only fair that I disclose a small detail - in case it looks as though I am suggesting anything along the lines of biomimicry. Diffusion is not the fastest way to know position, particularly when we are driving cars. It takes about an hour to establish a concentration gradient in a growing fly. But for the purposes of growth of a fly, this timeframe is adequate.
Of course, many questions arise. How sensitive is the "position reading" to variations in the diffusing molecule concentration? How sensitive is it to variation in the number of sensors on a cell to read concentration? How is it that nature pulls off this amazing feat of engineering, self-assembling a small fly using internal GPS systems, for millions of years with all its intricate details?
We don't have all the answers to this remarkable feat of engineering. In fact we know very little. But the glimpses we have into nature's algorithms of self-assembly using its GPS technology are intriguing and exciting. The animation [coming soon!] attempts to illustrate that. Note, that while we had the courage to ask only about self-assembly of a fly, if we figure out how a fly can be self-assembled, we would understand a lot about how we are self-assembled too. Nature's core algorithms seem to be shared across life, even if it gets tweaked a lot.
Lastly, life needs GPS not just for self-assembly. GPS is needed for daily activities too. For instance, when you have an injury, defender guard cells migrate to the site of injury to protect. The precision may not be as high as that needed for self-assembly though - they just need to make sure they head in the direction of the injury, they do not need to know the absolute distance to it.
Lastly, life needs GPS not just for self-assembly. GPS is needed for daily activities too. For instance, when you have an injury, defender guard cells migrate to the site of injury to protect. The precision may not be as high as that needed for self-assembly though - they just need to make sure they head in the direction of the injury, they do not need to know the absolute distance to it.
References
How cells know where they are - Arthur D Lander, Science 2013
Developmental pattern formation: insights from Physics and Biology - Anna Kicheva et al., Science 2012
Rise of the source-sink model - Alexander F Schier et al., Nature 2009
From Signals to Patterns: Space, Time and Mathematics in Developmental Biology - Julian Lewis, Science October 2008.
Robust formation of morphogen gradients - T BollenBach, Physics review letters 2008
